1: \begin{abstract}

2: We generalize the Euclidean $1$-center approximation algorithm of~\cite{coresets-2003} to arbitrary Riemannian geometries, and study the  corresponding convergence rate. 

3: We then show how to instantiate this generic algorithm to two particular settings: 

4: (1) the hyperbolic geometry, and 

5: (2) the Riemannian manifold of symmetric positive definite matrices.  

6: \end{abstract}

7: